The Big West’s Tragedy of the Commons

In 1969, UCSB ecologist Dr. Garrett Hardin published an essay entitled The Tragedy of the Commons in the journal Science which launched him into academic notoriety.  In essence, the essay illustrated the fact that when a shared and depletable resource is left unmanaged, the consumers of that resource act in their own short term interests and exhaust the resource, sometimes completely out of existence.  Commonly used examples to demonstrate the concept are shared grazing lands and shared fisheries.  Ranchers and fishermen understand they have competitors for the resource, and they anticipate that their competitors will consume it.  In order to ensure they enjoy the benefits of the resource themselves, they act in their own self interest and use the resource as much as they can while it still exists.  Failing to do so could drive them out of business.  Of course, if the resource can only regenerate at a certain rate and requires a certain baseline level to sustain itself, the resource will become depleted to the point where it can no longer regenerate itself.  For example, if fishermen take their stock below a certain level, there will be no more fish to breed.  Likewise, overgrazing can lead to the erosion of topsoil and thus result in barren lands devoid of vegetation.  So, in acting in one’s own short term self interest, one can do substantial harm to the shared resource which the collective relies on for long term success.

So, what can this possibly have to do with UCSB soccer?  It has to do with how the individual Big West schools schedule their out-of-conference (OOC) matches, and the impact it has on their collective RPIs.  Within the scheduling lies a paradox: in order to achieve a high RPI value, one must schedule teams that one anticipates will have relatively high winning percentages.  However, in scheduling those better-than-average teams, one is likely to see one’s own winning percentage suffer, and if every school in the conference pursues a challenging schedule, then no school will have a high opponents winning percentage (50% of the raw RPI) and thus no school can achieve a high value RPI as we all turn into the proverbial crabs in the bucket.  In pursing one’s own narrow self interests without consideration for the whole, it leads to collective demise.

While it is true that a challenging schedule is required for the possibility of earning an at-large bid to the NCAAs, Big West schools have seemingly been taking an individualistic approach in recent years.  When too many of them schedule with only its own self interest in mind, it virtually assures that all of them will suffer a similar RPI fate.  There is no better formula for achieving a high RPI than to maximize winning percentages within a conference through selective OOC scheduling while simultaneously minimizing the numbers of teams that act as an anchor that stunts the RPI potential of the others.

Coaches understandably answer the siren call to challenge their team by putting them on the road versus traditionally strong teams.  Steve Sampson has done just that by starting the 2017 season with 4 straight matches on the road versus Indiana, Notre Dame, Georgetown, and Maryland.  Sure, playing those matches tests one’s team, and if they can eek out a couple results, the RPI bonuses could be significant.  But much more likely, playing those teams will result in a poor to mediocre record entering conference play.  Of course, after Cal Poly lost to Notre Dame, Sampson was proud of his team.  A moral victory for him, but an RPI drag for everyone else.

When Cal Poly announced its schedule over the summer, it trumpeted the schedule it had put together.  If you’re a member of the Big West, especially in the North division, this boast caused alarm.

If the conference as a whole were to explore a mutually beneficial approach, it would likely raise everyone’s RPI.  Doing so would increase the odds of receiving multiple NCAA bids.  Receiving multiple NCAA bids increases the profile of the conference.  An increased conference profile helps recruit better players.  Better players results in better results, and better results yield better RPIs.  Before you know it, you’ve found yourself enjoying a beneficial positive feedback loop.

Of course, the conference doesn’t schedule the matches for its members.  The coaches do so, and those coaches have an obligation to a number of stakeholders to maximize their school’s individual success.  Maximizing success also ensures their own employment.  Without coordination with other coaches, they will do what is in their own best interest, and that likely means a challenging schedule that puts them in the best position to earn an NCAA bid.  Like fishermen and ranchers in an unregulated environment, they are out to extract whatever they can without thought about the long term consequences to the collective.

So, how should coaches approach scheduling?  A few ideas:

  1. Schedule as many matches at home as possible.  On average, one’s winning percentage is higher at home. Should you lose or draw to a team with a poor final RPI rank, the resulting penalty is largely self-contained to that school.  Plus, if you’re doing poorly at home versus poor teams, then you’re likely not good enough for the NCAAs anyway.
  2. Schedule as many teams as you can reasonably beat.  This doesn’t mean playing the worst teams you can find.  It means being realistic about your own quality and playing teams that are roughly equal or just below your expected quality.  Don’t worry about your opponents winning percentage!  If every Big West team follows this approach, most teams would realize a substantial RPI benefit once conference games are played, and you will have the opportunity to be in a favorable position when the NCAA Selection Committee makes its selections.  Granted, this scheduling approach is easier to implement in non-western states where the pickings are more plentiful, but beneficial match-ups do exist west of the Mississippi.  Pro tip: schedule average teams from very weak conferences.
  3. Schedule weaker teams from out-of-region.  Ideally, teams from the West should play the weakest teams from other parts of the country.  This approach captures RPI values that reverberate among all the Western teams.  Conversely, playing against the more prominent teams from the more prominent out-of-region conferences when one’s prospects for a result are poor endangers the RPI for all the teams that are reliant on a robust regional RPI framework.
  4. Schedule no more than 2 traditionally good teams.  Swinging for the fences like Cal Poly is doing this year just hurts the conference (as well as the entire region).  However, if your team is expected to peak in a given year, feel free to go for that one or two marquee victory on the road that can earn you a nice RPI bonus.  But be realistic!

Prior to the season starting and upon looking over the schedules of the Big West teams, I wrote about the concerns I had regarding our RPI potenital.  I know Vom Steeg understands the RPI quite well, and I suspect he shares my frustration with the rest of the conference, especially Cal Poly.  I also happen to know that former Gaucho and current Loyola head coach Neil Jones attempted to propose an RPI scheduling strategy among the head coaches of the Missouri Valley Conference.  I don’t know if that meeting of the minds prompted any coaches to change their approach to scheduling, but Jones’ attempt to seek a mutually beneficial strategy among the conference’s coaches was commendable and worth pursuing.

If these concepts were presented and if given the opportunity, would the coaches of the Big West employ a collective approach that would likely increase their prospects for success or might they reject cooperation (possibly through feigning cooperation) and prefer to employ a game theory approach where they attempt to anticipate the cooperative actions of others and then act in ways to solely enrich their own individual prospects?  Unfortunately, The Tragedy of the Commons suggests the latter approach as inevitable.

Match Days 1 and 2: Saint Mary’s and Siena

After our 2-1 exhibition loss to Westmont, there was much gnashing of teeth.  O-M-G, we haven’t lost to that team in 17 years, back when UCSB was still in the hinterlands of Division 1 soccer!  Something must be rotten in Denmark, and the downfall of the UCSB program is nigh!  Meh.  I didn’t watch the match, so I was spared the horrific spectacle one would expect from a 1 goal exhibition loss (/s).

So, let’s move on and discuss the results that matter and will actually bear some significance on our season.  Oh, but before I start, I wanted to mention that Arthur Wilkie from KCSB sports and I sat down in the KCSB studio on Saturday afternoon (between the Friday and Sunday matches) for a 50 minute chat.  A good deal of what I discuss below was also discussed with Arthur.  Anyway,  we were both in good spirits as his Atletico Madrid and my Dortmund had won earlier that morning.  Btw, you know someone knows and likes soccer when they root for a team other than Real Madrid, Barca, or from a wide range of EPL clubs.  Yes, mentioning that was also a shameless humble brag for myself.

Alright, Saint Mary’s.  This is roughly how we started the match, playing in a 3-5-2:

Dream Team 3-5-2 football formation

What is remarkable about the line-up is that it includes 5 freshmen including 4 pups with defensive responsibilities.  Now, Saint Mary’s fielded a very similar formation (albeit with much more experienced players) which contributed to an especially crowded midfield.  About midway through the first half, Vom Steeg abandoned the 3-5-2 and implemented a 4 man backline when he brought on Poulter for Acosta.  We then looked something like this and roughly stayed like this for the remainder of the match (with players obviously subbing in and out):

Dream Team 4-4-1-1 football formation

I don’t know exactly the reason, but we played many long, hopeful balls from the back that were dealt with predictably by Saint Marys’ tall center backs.  This was particularly true from the right side of our defense of Kashani (and later Gillingham) and Poulter.  Was it instructions from TVS?  Was it nerves from new players to not make a shorter pass and make a mistake that would lead to a goal?  Was it teammates not checking for the ball and providing a shorter, ball-possessing option?  It was probably a combination of these factors.  I know that Kashani made a bad pass very early in the match which led to a St Mary’s opportunity, and maybe after that, it was a “safety first” approach.  But whatever the case, those long balls disappeared into a black hole.  So, the game never resembled much of a well-played soccer match so I’m not going to analyze it much more.  Plus, our approach versus Siena was worlds different from Saint Mary’s, so let me pause to discuss some individual impressions that were left on me:

Kashani knows how to tackle, gets stuck in, and generally plays a physical but clean defensive game.  Between he and Mendoza on the left, our wing defense is going to be very strong.  However, his passing leaves something to be desired (as did his crossing versus Siena).  But it’s much, much easier to clean up those parts of someone’s game than it is to teach someone defensive intuition which he clearly has.  He won virtually every tackle and duel over both matches.

Carillo was solid versus Saint Mary’s, making one tremendous fingertip save over the bar after a shot deflected off Acosta.  Earlier in the match, he correctly came off his line to thwart a shot inside the 6 yard box.  He commanded his box and communicated well to organize his defense.  On several occasions, when he had the ball at his feet, he allowed an opposing player to close too much on him.  Some of his clearances were a bit low, and he played one ball straight to a St. Mary’s player just outside the box.  But overall, he did quite well, and it’s hard to complain about 220 scoreless minutes.

In my mind, Adames has locked up one of the center back starting roles.  He has a good combination of size, strength and speed, and he has a clear knack for defending, showing good decision making, being powerful in the air, and effective 1-v-1 defending.  It’s when he’s on the ball that he can be of two minds, allowing an attacker to capitalize on his indecision.  Like Kashani’s weaknesses, this can be corrected as he becomes more comfortable with more playing time.  The question for me will be who will be partnered with Adames.  Versus Saint Mary’s, it was primarily Poulter.  Versus Siena, Salgado was paired with Adames, likely to ensure a more patient build-up from the back.

Since I’ve started bringing up the Siena match, let me illustrate how we lined up for that fixture:

Dream Team 4-2-1-3 football formation

As I mentioned, Salgado moved back to CB with Ilskens taking the holding/defensive midfield spot.  Michael played wider this match which suits his ability to beat players 1-v-1.  Playing more centrally in the prior match caused him to be more prone to be bottled up and dispossessed by multiple defenders, so I think he will (and did) have much more success out wide versus Siena.  With older and more composed players like Salgado and Ilskens, our buildup play from the back was much improved.  Additionally, the communication on the field was very good, letting players know what was around them and what options they had as they received the ball.  Of course, we were also playing a different team that were playing us differently, but Salgado and Ilskens were positive influences nonetheless, and their lead showed younger players how they were expected to play.   As an aside, I see the impact of expectations all the time in pick-up soccer.  I play with one group of players with personalities that like to dribble the ball, try killer balls, and generally not pass or possess much as a team.  That’s the culture of that group, and if you try to keep anyone accountable for their risky/panicked/selfish play, your message isn’t well received.  When I play with another pick-up group, you’re expected to possess the ball and keep the ball on the ground.  If you play a risky ball rather than a safe ball (even if it comes off well!), you start to hear it from your teammates pretty quickly.  So, after the St. Mary’s match where we didn’t possess much, and didn’t use the width of our field, bringing in more experienced players versus Siena in order to keep possession and build more patiently was the right move by TVS.  It resulted in a number of runs up the wings, crosses into the box, and corner kicks.  Unfortunately, we got away from this a little bit in the second half.

I am confident that TVS is dialing in on playing the right players in the right positions, and the starting line-up versus Siena looks about right for me.  Tellechea up top is good.  Billingsley behind him allows him to continue to attack but also takes better advantage of his high work rate and ability to win the ball.  I might prefer him more at left wing, but behind the striker works well.  Ilskens at holding midfield works very well.  If a dependable CB is identified to pair with Adames, Ilskens and Salgado as shielding midfielders might also work.

While we didn’t manage to score a goal over our first two matches, we also haven’t conceded a goal, which is remarkable given how new and young our defense is.  We also played much, much better versus Siena, and really deserved a better result, but there are definite signs of improvement.  The goals will come.  We know that Carillo, Kashani, Adames, and Michael will be key contributors, in particular Michael.  Tellechea will also likely figure into whatever success we manage this year.  Additionally, Ammer will be available for selection versus St. John’s as he has had to sit per NCAA eligibility rules.  Also, it looks as though he will be reclassified as sophomore rather than his current status as a junior.  Axel Mendez, former Big West Freshman of the Year, remains more of a mystery.  He is with the team, but I have no information regarding his eligibility.  Both players would inject significant playmaking into the team.

Now that I’ve provided some hope and optimism, it’s time for a bit of a reality check.  As I forewarned in my Rooting Interests for 2017, our RPI worksheet shows clouds on the horizon.  The Big West is a collective 3-8.  But it gets worse!  Our Big West North brethren are 0-4.  Adding up all the wins/losses for our opponents amounts to an overall opponents record of 9-19.  I predict that we will need to win the Big West Tournament in order to qualify for the NCAAs.  The bright spot in this is that it frees us from attaching too much significance from our out-of-conference results.  As long as we continue to improve and show team cohesion and enter Big West play as a strong team, we should be in decent shape to contend for the NCAA auto-bid and enter that tournament with enough momentum to surprise some teams.

Next up is a match at USD on Friday which is coming off two losses versus Indiana and Notre Dame during a road trip to the Hoosier State.


Rooting Interests for 2017

If you’ve followed my blog over the last year, you know that the tournament seeding and at-large selections are almost entirely dictated by a team’s RPI rank.  While rooting for the Gauchos and celebrating their wins is what makes following the team compelling, in the grand scheme of things, our own results aren’t actually the biggest determiner of our post season prospects, at least not in the 18 game regular season.  Rather, the regular season’s only importance is to qualify for the NCAA tournament and then position ourselves within the tourney, and then hopefully find some magic and go on a run. Since our all-time NCAA tourney record is 12-2-0 at home and 6-8-2 away from home, earning a top 16 seed should be the primary objective every year.  Anything less than a seed makes it an uphill climb to the College Cup.  Then again, 2006.

The secret to a successful season is essentially 3 fold.

First, schedule teams you can beat that you believe will have decent records.  Playing (and beating) mediocre teams from a weak conference represents the ideal opponent.  Obviously, scheduling the right opponents and being able to accurately predict their approximate winning percentage is an inexact science.

Second, hope that your conference (and particularly the Big West North) do well in their out-of-conference (OOC) matches.

Third, take care of your own business while maximizing RPI bonuses and minimizing RPI penalties.

For those new to the RPI (Ratings Percentage Index), it is calculated as follows:

25% is one’s own winning percentage

50% is the winning percentage of one’s opponents

25% is the winning percentage of one’s opponent’s opponents

The above calculations result in the raw RPI.

Additionally, bonus points are awarded for wins or draws for away/neutral site games versus a team with a high RPI rank.  Factoring in bonuses and penalties yields the adjusted RPI which represents the final calculation that the NCAA uses.  When I refer to the RPI, I mean the adjusted RPI.  A win at a top 15 team yields the highest possible bonus where a draw at a neutral site versus a 46-60 team yields the smallest bonus.  Likewise, a loss at home versus a low ranked RPI team imposes a penalty.  Bonuses and penalties can be the difference between a better/worse seeding or even earning an at-large berth into the tournament.  For an entire list of bonuses/penalties, check out this document.

Back to the 25/50/25 calculations.  By far the biggest determiner of a team’s RPI value is the cumulative opponent’s winning percentage (OWP).  Once the schedule has been set and the games start being played, this value is completely out of our control.  If you’re in a conference with teams that aren’t winning, no matter how stellar your own winning percentage, your RPI ranking will be mediocre even if you finish the season 20-0-0.

Let’s have a quick look at the opponent’s opponent’s winning percentage (OOWP).  This calculation incorporates many results, and thus this value tends to the mean (.5000).  For example, while one’s OWP is the aggregate of around 20 values, the OOWP is the aggregate of around 400 values (20 x 20).  For this reason, the OOWP value difference from one team to the next is not very large and thus its not a large differentiator of one’s overall RPI calculation.

So, what matters most is the OWP followed by one’s own winning percentage.  If both of these values are north of .5000, you’re already looking pretty good.  However, if the OWP is below .5000, one’s potential RPI ranking is going to have low potential.

I should pause here and clear up one misconception that I come across a lot.  Most people focus on an opponent’s RPI rank rather than an opponent’s winning percentage.  The rank only matters when there is a bonus or penalty in play.  What really matters is the opponent’s winning percentage.  For example, there is much greater value in playing a 14-6 team with an RPI rank of 110 than a 10-10 team with an RPI rank of 20.

By far the most important part of the season is the OOC portion.  This is where our RPI die will be cast.  Why?  Because once we hit our conference schedule, the winning percentages of 10 of our 18 matches becomes a more-or-less closed system.  In other words, the average winning percentage of all our future opponents will be set since we all start playing each other. Of course, at the very start of conference play we want the North teams to beat the South teams, but for the most part, our end-of-the-season OWP and OOWP values will largely be set in stone at the start of conference play.  Ideally, one of the North teams or one of our away South opponents will have a high rank to provide us a potential RPI bonus.  However, on the flip side, a home opponent with a very low RPI rank can imperil our RPI since losing to or drawing that team at home will incur a penalty, and when it comes to rivalry matches, anything can happen.

Last year, it was clear that when we ended our OOC schedule, we would need to win the Big West tournament in order to qualify for the NCAA tournament due to how mediocre our conference fared during the OOC portion of the schedule.  This was already known despite the fact that we had played only half our matches.

Historically, out of the 24 total conferences, the Big West has been everything from stellar (as high as #2 or #3 and earning 4 bids to the NCAAs) to very mediocre (#13 in 2016 with one bid).  Of course, UCSB is typically the best team in the conference (or certainly top 2), so the rest of the conference drags us down while we do the heavy lifting to elevate the conference.  How the Big West conference does year-to-year is a crap shoot.

By far the most important item to track this year is the UCSB RPI worksheet.  Bookmark it now (or visit my site and find the link on the right side).  Our worksheet is part of a website created and maintained by a UCSB fan, and he has created a worksheet for every NCAA D1 team.  He also provides other useful pages including a daily scoreboard, complete RPI ranking, conference standings, conference RPI ranking, and a tournament projection.

With the RPI primer out of the way, let’s have a look at what UCSB fans should be rooting for this year, in order of importance.  Each team’s 2016 record and RPI rank is in parentheses.

1.  Big West North

We have 18 matches on our schedule, not counting the BW tournament.  Six of those 18 matches, a full third, are against our divisional foes.  We also play each team once away, so if they have an RPI ranking of 75 or better, a bonus is in play.  However, if their RPI rank is very low and we lose to them at home, we incur an RPI penalty.  More than anything else, we want Cal Poly (5-9-3, #116), Sac State (9-8-3, #100), and Davis (8-9-3, #112) to have high winning percentages.  There is already cause for concern since Poly bragged about having the top schedule in the country.  This spells big trouble for us as their winning percentage prior to conference play will likely be poor.

Jul 13 

With 26d till arrive 4 2017 season, our captain reveals 8 tournament teams top schedule in the country!

If we have just one anchor in our division, the other 2 schools simply won’t be good enough to neutralize the heavy drag that one team with a horrible WP will have on us.  It’s highly possible that Cal Poly has predestined itself for being our division’s anchor.

2.  Big West South

We have 4 matches versus these schools.  In addition to that, so do our divisional rivals, which has an amplifying impact on the OOWP.  In particular, we want UCI (5-13-1, #140) and CSUN (9-7-5, #74) to have high RPI ranks since we play them on the road, and we can earn bonus points with a result, but we also want CSUF (10-9-4, #102) and UCR (8-10-2, #133) to do well.  Go Eaters, Matadors, Titans, and Highlanders!

3. West Coast Conference

We have 4 matches versus this conference as we play USD (6-9-3, #89) , Saint Mary’s (8-8-3, #62), Portland (12-5-2, #43), and Pacific (13-4-2, #29).  So, we want those teams to do well, but we also want to root for their conference mates unless they’re playing a Big West team.  Like us, Davis also plays Saint Mary’s and USD (and also LMU, USF).  Cal Poly plays Saint Mary’s as well as Portland while Sac State plays USD, Saint Mary’s, Pacific in addition to LMU.   We play USD in San Diego, so we want USD to do particularly well for a potential bonus.  Given the number of matches we play versus the WCC and given the number of matches our fellow North teams play (including many common opponents), the WCC is almost as important as the BW South.  So, in a nutshell, we want USD, Saint Mary’s, Portland and UOP to do particularly well, but we also want the entire conference to do well against any non-BW team.  Go Toreros, Gaels, Pilots, and Tigers, and the rest of the WCC!

4.  San Diego State (9-4-6, #30)

In reality, the next several rooting interests are on par with each other.  San Diego State gets the nod as we play at SD, so there is a potential RPI bonus.  Also, Cal Poly plays SD State, so it’s in our interest that SD State does well.  In normal years, we might also be reluctantly rooting for the Pac-6 conference, but the BW North is playing very few matches against that conference.  In fact, other than UCSB and Poly playing SD State, no other BW North team has a match against the Pac-6.

5.  St. John’s (6-7-4, #125)

We play at St. Johns.  Go Friars, and by extension, Big East!

6.  Akron (13-6-3, #27)

We host Akron.  Go Zips and Mid-American Conference!

7.  Siena (8-8-5, #135)

We host Siena.  Go Saints and the MAAC!

8.  West versus the Rest

It’s better for UCSB if teams from the West beat teams from the South, Midwest, or East in order to capture RPI values to our region and have those values reverberate among common opponents.  It’s very good when teams from a West conference like the WAC beat out-of-region teams even if we don’t play any of the teams from that conference.  If one region wins cumulatively more games over another region, this small difference can have a very large impact on RPI.  Always root for the teams from Western conferences over any other part of the country.  Go West!

9.  Draws

Generally, if you’re not actively rooting for a team to win based on the criteria listed above, then root for draws as it pulls teams closer to .5000.  The more regression to the mean of other schools/conferences, the better it is for us to earn a high seed or an at-large bid.  With approximately 206 Division 1 teams, you generally need an RPI rank of about 40 for at-large consideration.  Draws pull a team toward the middle of the pack, but you need to be in the top 20% of RPI for an at-large bid.   Go draws!

10.  Losses for High RPI Teams

Later in the season, we generally want higher ranked teams to lose to lower ranked teams.  That is certainly the case where an out-of-contention team plays a top 60 team, but there are a few caveats here.  For example, if a top 5 team is playing the #30 team, we want the #5 team to win IF we’re not in contention with the #5 team.  Also, one must consider RPI bonuses, so if the #20 team is hosting the #45 team, we would probably want the #20 to win so the #45 team doesn’t jump up in the RPI standings.  Anyway, losses by highly ranked RPI teams is highly situational and becomes increasingly relevant as our own situation becomes more clear in the latter half of the year.

11.  Losses for Big Conferences

Some conferences are traditionally strong.  Generally speaking, we want teams from weaker conferences to beat teams from stronger conferences.  Otherwise, the stronger conferences will dominate the NCAA seeding.  For example, last year the ACC ended the year with the #2, #3, #6, #9, #10, #13, #17, and #20 teams.  That’s because the ACC dominates its OOC schedule and thus captures RPI values for their conference.  Yes, teams from the ACC are very good, but they also benefit from the fact that they have so many scheduling options on the East Coast while also having the gravitas to force teams to play games at ACC schools.  Yes, the RPI is heavily flawed and can be gamed.  Just know this… the ACC is evil and we want them to lose!

With all this RPI talk, I want to make it clear that I actually loathe its use.  I’m not necessarily opposed to the NCAA using a statistical model for determining the relative performance of each team, but the RPI is a poor tool.  There are much better models that the NCAA could use, but we’re stuck with this one, so it’s important to know how it works.

When you see polls cited by the NCAA or UCSB’s very own public information office in match previews, those polls are meaningless.  In fact, citing polls promotes ignorance since the polls mislead people into believing they matter.  I suppose one can pick the poll where one’s team has the highest ranking and use it for promotional or recruiting purposes, but it’s important for my readers to know that they have zero/zilch/nada relevance when it comes to our prospects for the NCAA tournament, and in the end, that’s all that matters.

To sum this all up, the regular season’s purpose is to make it into the NCAA tournament and compete for a national title.  In order to position oneself favorably with a bye and home field advantage, one must earn a high RPI ranking.  In order to earn a high RPI ranking, one needs one’s opponents to have a high winning percentage.  Winning one’s own games certainly helps.

If you have a question about the RPI, please leave a comment below.